MDS Associates, Limited Partnership v. United States

37 Fed. Cl. 611, 1997 U.S. Claims LEXIS 50, 1997 WL 112609

• Court: United States Court of Federal Claims
• Decided: March 11, 1997
• Docket: No. 93-429C
• Status: Published

Opinion

OPINION

MILLER, Judge.

This case is before the court after trial. MDS Associates, Limited Partnership, and F. Gregg Bemis, Jr. (“plaintiffs”), brought suit against the United States pursuant to 28 U.S.C. § 1498 (1994), for reasonable and entire compensation for the unauthorized use and manufacture of the invention described in United States Patent No. 3,737,902 (the “ ’902 patent”). By agreement of the parties, only the liability portion of the case was tried, with damages to be determined at a later date if the court found defendant liable for infringement of the ’902 patent. The scope of trial was limited further by the parties’ August 4, 1995 stipulation for trying the case and binding the parties based on representative models of accused systems.¹

FACTS

1. Automated collision avoidance

The technology at issue in this case is automated collision avoidance radar systems. One way to prevent the collision of two objects is to predict the closest point of approach (the “CPA”). The CPA between two objects is the point at which the two objects come closest to one another. Although CPA can be used to prevent the collision of any two objects, the technology involved in this case was developed primarily to prevent the collision of two ships. The ship performing the CPA calculation is called “ownship” and the second ship, .whieh ownship is trying to avoid, is called the “target.”

The relationship between ownship and the target at CPA may be expressed in several different ways. “CPA Range” refers to the distance between ownship and the target at the closest point of approach. “CPA Bearing” refers to the direction of the target in relation to ownship at the closest point of approach. “CPA Time” refers to the time at whieh ownship and the target will be closest to one another.

Mariners have been predicting CPA during most of the Twentieth Century. Prior to the invention of automated collision avoidance systems, they performed the CPA calculation manually using a plotting chart known as a maneuvering board. A mariner would plot the positions of the target at at least two points in time on the maneuvering board. By continuing the line formed by connecting the known positions of the target, the future position of the target could be predicted. Assuming the target did not change course or speed, the target would always be found along this extended line, known as the relative motion line. Having found the relative motion line, a mariner could determine CPA Range and CPA Bearing by dropping a perpendicular from own-ship to the relative motion line. The length of the perpendicular represented CPA Range, and the angle of the perpendicular, CPA Bearing.

With the invention of radar, mariners began to automate the process of predicting CPA. The radar identified the position of the target, and then a mariner would plot the target position on a maneuvering board. Although certainly a useful navigational aide, predicting CPA with a maneuvering board was a time-consuming process. If either ownship or the target changed its course or speed, a mariner had to begin the maneuvering board calculus all over again.

Automated collision avoidance relieves a mariner of the need to use a maneuvering board when predicting CPA. By connecting a radar system, a compass, and a speed log to a computer, the computer receives information about ownship and the target. The computer then processes this data and provides a mariner with a predicted CPA With the touch of a few buttons, the mariner now can access information that previously required a series of mathematical calculations.

2. Development of the patented invention

The ’902 patent lists 4 inventors — Robert M. O’Hagan, Richard H. Sorensen, Joseph F. DeSpautz, and James E. Carroll, Jr.² The inventors developed the system covered by the ’902 patent while employed at Ma-riñe Digital Systems, Ine. (“Marine Digital”), a corporation formed by Messrs. O’Hagan, Sorensen, and DeSpautz for the express purpose of developing and marketing an automated collision avoidance system. The development and marketing of the system produced by Marine Digital was backed financially by F. Gregg Bemis, Jr., an investor, who testified that he lost $600,000.00 on the project.

On August 26, 1969, the inventors filed a document entitled “Disclosure Document” with the United States Patent and Trademark Office (the “PTO”). The Disclosure Document described a computer automated collision avoidance apparatus. On August 19, 1970, the inventors filed an application for a patent with the PTO. The PTO initially rejected the application. However, after the inventors submitted an amendment to the initial application, the PTO issued the ’902 patent on June 5,1973.

During the past 27 years, the ’902 patent has been owned by a number of entities. See MDS Associates, Limited Partnership, et at. v. United States, 31 Fed.Cl. 389, 390-91 (1994) (denying motion for partial summary judgment based on real party in interest and discussing chain of ownership of ’902 patent). On December 16, 1970, shortly after submitting the patent application to the PTO, the inventors assigned the patent application to Marine Digital. Marine Digital, in turn, assigned its rights in the patent application to State Street Bank & Trust Company as collateral for a $250,000.00 loan secured by the Small Business Administration (the “SBA”). Later Marine Digital became insolvent, defaulting on its SBA-secured loan, leaving a remaining balance of $193,533.00. State Street Bank & Trust Company then assigned its rights in the patent application and resulting patent to the SBA effective on July 17, 1972, and March 5, 1981. On March 11, 1981, the SBA assigned the ’902 patent to Mr. Bemis for consideration of $1,000.00 and one third “future net receipts,” including “earnings, royalties, litigation recoveries and all other income emanating from the ownership and use of the Patent until SBA is paid $270,000 or until the Patent expires, whichever comes sooner.”

On August 29, 1986, Mr. Bemis entered into an exclusive license agreement with Mr. O’Hagan whereby Mr. O’Hagan agreed to pay Mr. Bemis 51% of the “Gross Receipts” received by Mr. O’Hagan or any sublicensee from the ’902 patent; the agreement was subject to the SBA’s above-mentioned rights. Paragraph 6 of this agreement specifically assigned to Mr. O’Hagan the right to file suit for recovery of damages and profits related to past or future infringement of the ’902 patent. Following the Bemis/O’Hagan license agreement, Mr. Bemis assigned his rights in the ’902 patent to MDS Patent Corp., a Massachusetts corporation of which Mr. Bemis was the sole director and shareholder. Mr. O’Hagan likewise assigned his rights as a licensee to plaintiff MDS Associates Limited Partnership, of which Messrs. O’Hagan and Sorensen are partners. MDS Patent Corp. was dissolved by the State of Massachusetts on December 31, 1990, with the result that Mr. Bemis was once again the owner of the ’902 patent. At the date of the filing of this lawsuit, Mr. Bemis owned the ’902 patent, subject to the exclusive license of MDS Associates, Limited Partnership, to file suit for recovery of damages and profits related to past or future infringement of the ’902 patent.

Sometime in late 1970, the inventors, working at Marine Digital, produced the first working prototype of the system disclosed in the ’902 patent application. The system was called COMAN, an acronym whose meaning none of the inventors can remember. Marine Digital constructed a mocked-up ship’s bridge in a building facing Plymouth Harbor in order to allow potential purchasers to test COMAN’s collision avoidance function on ships afloat in the harbor. Prior to going out of business, Marine Digital sold six COMAN systems.

3. The accused representative systems

Plaintiffs originally identified five systems that they believed infringed the ’902 patent. During the course of discovery, the parties recognized that there were multiple versions of many of the accused systems. In order to provide for a manageable trial, the parties entered into a joint stipulation wherein the parties designated a representative model for each accused system. The representative models agreed to by the parties are, as follows:

[[Image here]]

MDS Associates, Limited Partnership, et al. v. United States., No. 93-429C (Fed.Cl. Aug. 4, 1995) (Joint Stipulation on Representative Models of Accused Systems). Each is used by the United States Department of the Navy, except COMDAC, which is a United States Coast Guard system.

The NTDS, AEGIS, HYCATS, and COM-DAC systems commonly are referred to as command data systems. At the core of these four systems is a digital computer that receives and processes data signals received from a number of sensors, including a radar, sonar, and compass. Although these systems are capable of predicting CPA data, they also perform an extensive array of other functions, such as tracking air, surface, and subsurface targets; assessing threats; targeting and firing weapons; navigating the ship; and exchanging tactical data with other ships in the fleet. The AN/SPA-25G, unlike the other accused representative systems, is not a tactical command data system. Rather, it is a stand-alone radar display containing a digital computer that can compute CPA data.

DISCUSSION

I. Infringement

Whether an accused device infringes a patent is a question of fact.³ Braun, Inc. v. Dynamics Corp. of America, 975 F.2d 815, 819 (Fed.Cir.1992). To determine infringement, the court employs a two-step analysis: “ ‘First, the claim must be properly construed to determine its scope and meaning. Second, the claim as properly construed must be compared to the accused device or process.’ ” Ekchian v. Home Depot, Inc., 104 F.3d 1299, 1302 (Fed.Cir.1997) (quoting Carroll Touch, Inc. v. Electro Mechanical Sys., Inc., 15 F.3d 1573, 1576 (Fed.Cir.1993)).

1. Constmction of the claim

The ’902 patent contains 14 claims. The parties agree that Claim 1 of the ’902 patent is an independent claim upon which all of the remaining claims are directly or indirectly based. For purposes of determining infringement, the court is required to interpret only Claim 1. If Claim 1 is infringed, all subsequent dependent claims are also infringed. Likewise, if Claim 1 is not infringed, none of the subsequent allegedly infringing dependent claims is infringed.

Claim 1 states:

Collision avoidance apparatus comprising,

[a] a source of a signal representative of the velocity of a prime vehicle in a predetermined fixed coordinate system,

[b] means for providing a target position signal representative of the position of a target object relative to that of said prime vehicle,

[c] means for providing a target velocity signal representative of the velocity of said target object in said fixed coordinate system,

[d] and means responsive to the velocity representative and target position signals for providing a vector signal representative of the distance and bearing between said prime vehicle and said target object at the closest point of approach between said prime vehicle and target object were the prime vehicle velocity and target object velocity to maintain the values characterized by the representative signals until said closest point of approach oecurs.[⁴]

Claim 3 states:

Collision avoidance apparatus in accordance with claim 1 wherein said signal representative of velocity of the prime ve-hide is representative of a proposed velocity for said prime vehicle.

Claim 11 states:

Collision avoidance apparatus in accordance with claim 1 and further comprising means responsive to the velocity representative and target representative and target position signals for providing a time signal representative of the time from the present to the time of said closest point of approach were the prime vehicle velocity and target object velocity to maintain the values characterized by the representative signals until said closest point of approach occurs.

Claim 12 states:

Collision avoidance apparatus in accordance with claim 1 and further comprising means responsive to said vector signal for providing a display of said distance and bearing in at least one of alphanumeric values in list form and line vectors.

The key to the proper construction of Claim 1 lies in the patent’s use of means-plus-function language in limitations (b)-(d). 35 U.S.C. § 112 (1994), permits the use of mean-plus-function language:

An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.

35 U.S.C. § 112 para. 6. While section 112 clearly permits the use of means-plus-function language, a means does not cover the entire possible universe of means for performing the specified function. Laitram Corp. v. Rexnord, Inc., 939 F.2d 1533, 1536 (Fed.Cir.1991). Rather, “the court construes the means-plus-function language to encompass the structure, material, or acts described in the specification and the equivalents thereof.” Sofamor Danek Group, Inc. v. DePuy-Motech, Inc., 74 F.3d 1216, 1220 (Fed.Cir.1996).

The parties agree that Claim 1(d) sets forth a collision avoidance system that calculates a CPA vector in response to three specifically defined input signals. Claim 1(a) defines the first input signal, ownship velocity. Claim 1(b) defines the second input signal, target relative position. Claim 1(c) defines the third input signal, target velocity.

The parties do not dispute the nature of the function claimed by the ’902 patent, but they disagree as to the scope of the means-plus-funetion limitations. Defendant takes the position that the ’902 specification fails to disclose a means for performing the function claimed in Claim 1(d), namely the calculation of a CPA vector using an ownship velocity input signal, a target position input signal, and a target velocity input signal.⁵ According to defendant, the ’902 specification discloses a means for calculating a CPA vector using only two target position input signals.⁶ Plaintiffs counter that the ’902 specification does disclose a means for calculating a CPA vector based on an ownship velocity input signal, a target position input signal, and a target velocity input signal.

The parties’ disagreement over the construction of Claim 1 is best illustrated through defendant’s defense of non-infringement. Defendant begins from the premise that the means for performing the data processing function described in Claim 1 are determined by the operating program contained in the system’s computer. The operating program described in the ’902 specification consists of several mathematical equations that can be programmed into a computer. Defendant argues that the equations for the software program disclosed in the ’902 specification do not employ the ownship velocity and target velocity signals that are claimed in Claim 1. Instead, the equations use two target position fixes to determine CPA Range and CPA Bearing.

William M. Pease served as defendant’s expert in the field of automated marine collision avoidance engineering. From 1970 until 1992, as Director of Digital Technology and Corporate Engineering Staff at the Raytheon Company, he was involved intimately with the design of three generations of collision avoidance systems — RAYCAS, RAYPATH, and an unnamed system built into the Pathfinder ST radar.⁷

Mr. Pease gave defendant’s interpretation of Claim 1 through the use of two exhibits, DX 5 and DX 8. DX 5 graphically depicts how one can determine CPA Range and CPA Bearing by using two target relative position fixes, the method of calculation used by early mariners and the method that defendant argues is disclosed in the ’902 specification. See swpra p. 2 (discussing use of maneuvering board to calculate CPA Range and CPA Bearing).

[[Image here]]

DX 8 represents the same encounter between ownship and the target as does DX 5. However, in DX 8, the position of the target is only known at one point in time. To establish the relative motion line necessary for predicting CPA, DX 8 uses ownship and target true velocity, rather than a second target position. The subtraction of the velocity vector of ownship from the velocity vector of the target produces a vector that falls on the relative motion line. Once the relative motion line is determined by vector subtraction, one can predict CPA by dropping a perpendicular from ownship to the relative motion line, as in DX 5.

Defendant argues that the diagrams and equations set out in the ’902 specification do not employ ownship or target true velocity inputs to calculate CPA Range as shown in DX 8. Instead, the equations employ two target position fixes, as shown in DX 5:

Q [By Defense Counsel] Does the program described in the ’902 specification in any way use a prime vehicle true velocity signal as defined in Claim 1, Limitation A, to calculate the distance portion of the CPA vector signal?

A [By Mr. Pease] No.

Q Does the program described in the ’902 specification in any way use a target true velocity signal as defined in Claim 1, Limitation C to calculate the distance portion of the CPA vector?

A No.

Q Does the CPA program that is described in the ’902 patent specification correspond with any of Defendant’s exhibits that you have testified about today for predicting CPA?

A Yes, Exhibit 5.

When testifying about his interpretation of Claim 1, Mr. Pease conceded that the ’902 specification contains equations for determining ownship and target velocity in Equations 1-7 and Figure 2A of the patent. However, Mr. Pease was of the opinion that, despite the presence of these equations in the ’902 specification, the ’902 patent does not use ownship and target trae velocity to compute CPA Range and CPA Bearing. In other words, Mr. Pease explained that the specification provides equations for calculating own-ship and target true velocity, but that the system described in the ’902 specification does not employ the result of these calculations in predicting CPA Range and CPA Bearing:

Q [By Defense Counsel] Mr. Pease, would you like to explain why the velocity signals depicted in Figure 2A are not used in the calculation of CPA distance?

A We’ve been discussing how the small Vector K is exactly the same as a vector we can construct with the vectors that appear in Figure 2A. That’s a relationship, however, in the equations of the patent, that same Vector K is obtained from the two radar fixes, only. That’s Equations eight through 11.

So, in effect, we abandoned the development of K from figure 2A and proceeded to use the two radar fixes to compute K.

Mr. Pease also explained why the ’902 specification contains equations for calculating ownship and target velocity although the velocity calculations are not used to predict CPA. The computer calculates velocity as an aid to a mariner; once the velocity is computed, it is transmitted to a display monitor. While a mariner may find the velocity information useful, it is unrelated to the means by which the computer predicts CPA Range and CPA Bearing. If one simply deleted Equations 1-7 from the patent, the computer would still be capable of predicting CPA Range and CPA Bearing.

One of the co-inventors of the ’902 patent, Mr. Sorensen, testified as plaintiffs’ expert witness. Trained as an electrical engineer, Mr. Sorensen has been involved with radar since 1951 and has extensive experience in the maritime application of digital computers. He disagreed with Mr. Pease’s interpretation of the ’902 patent. Mr. Sorensen first identified the equations for calculating ownship and target true velocity in the ’902 specification. According to Mr. Sorensen, Equations 1-4 of the ’902 specification contain instructions for calculating ownship and target true velocity. The ownship and target velocity vectors are also depicted in Figure 2A of the ’902 specification. Line 25 of Figure 2A represents the ownship velocity vector, and Line 27 of Figure 2A represents the target velocity vector:

Q [By Plaintiffs’ Counsel] Mr. Sorensen, drawing your attention to Figure 2A of the 902 patent, can you identify a vector that is representative of the velocity of a prime vehicle in the system as claimed in subpar-agraph 1A?

A [By Mr. Sorensen] That is the vector that’s between the position PI and the position P2, whose names are one.

Q Is that this line?

A The line which is 25 on the drawing. And, it’s given information.

Q Mr. Sorensen, again referring to Figure 2A, can you indicate where the target velocity signal described in subparagraph 1C of Claim 1 can be found?

A Yes, that is the line which is 27. There is a vector between the point Pl-prime and P2-prime, which is a calculation within the Means 20 computer. And, it is the velocity vector, since the distance divided by the known time is equal to the velocity.

Mr. Sorensen then explained that once ownship and target true velocity are determined in Equations 1-4 of the ’902 specification, Equations 5, 6, and 7 convert true velocity into relative velocity, a conversion represented by the change from Figure 2A to Figure 2B:

Q [By Defense Counsel] Is that true target velocity signal used in calculating CPA parameters?

A [By Mr. Sorensen] It is in that we don’t know the relative velocity until we subtract away ownship’s velocity. The only time that relative velocity would mean anything without ownship’s velocity is if ownship wasn’t moving.

So, we deemed the true velocity, in order to be able to find out what the relative velocity is.

Q From the relative velocity, you can determine the relative movement line in Figure 2B?

A Yes, by using the same procedure that’s shown over here in Defendant’s Exhibit 8, where ownship’s velocity was subtracted from the target vessel’s true velocity-

Q, Is that process of converting from the trae motion of Figure 2A to the moving coordinate system or relative movement described—

A That’s exactly—

Q —in the 902 patent?

A That is exactly the process.

Q Where is that described in the 902 patent, in the mathematics?

A The mathematics are described by Equations 6 and 7 of the actual transformation of axis, and it’s vectorially explained, described, I should say.

For a number of reasons, the court finds that the means set forth in the ’902 specification do not employ ownship and target velocity to predict CPA Range and CPA Bearing. First and foremost, Mr. Pease’s expert testimony, including his use of DX 5 and DX 8, was extremely clear and consistent. Mr. Pease explained that although the ’902 specification did set up equations for determining ownship and target true velocity, the equations did not use ownship and target velocity to calculate CPA Range and CPA Bearing. In support of his interpretation of the ’902 specification, Mr. Pease discussed DX 12, a diagram depicting the inputs that are used in the ’902 specification to predict CPA. Of particular importance to defendant’s case, DX 12 also correlates each input with an equation in the ’902 specification. By correlating each input with an equation, Mr. Pease demonstrated how the equations in the ’902 specification direct the computer to predict CPA Range and CPA Bearing in a manner resembling DX 5 (two target position inputs), not DX 8 (ownship and target velocity inputs).

Mr. Sorensen’s testimony, on the other hand, was undermined by prior inconsistent testimony given at two earlier depositions.⁸ Defendant deposed Mr. Sorensen twice, once on October 19-20, 1995, and again on May 20-21, 1996. During the earlier deposition, Mr. Sorensen testified that ownship and target true velocity were not used in the ’902 patent to calculate CPA Range and CPA Bearing. However, during the later deposition, Mr. Sorensen changed his testimony, testifying that the equations in the ’902 specification in fact did employ ownship and target true velocity to calculate CPA Range and CPA Bearing.⁹ The catalyzing event was that, in the interval between the depositions, Mr. Sorensen took on the role of plaintiffs’ expert for trial. He explained that, as a consequence, he analyzed the technical documentation for the accused representative systems and spent additional time reviewing the ’902 patent. As discussed infra at 17, the accused representative systems use own-ship and target velocity, not two target position fixes, to calculate CPA. If Mr. Sorensen had not changed his testimony, his own opinion would have mandated a finding that the accused representative systems did not infringe the ’902 patent. While the court certainly does not attribute any dishonesty to Mr. Sorensen, it cannot help but find that Mr. Sorensen’s inspection of the accused technical documents, in combination with his prominent role in this case, affected his opinions.

Mr. DeSpautz, one of the co-inventors and the chief mathematician for Marine Digital, also testified that the equations in the ’902 specification use ownship and target velocity to calculate CPA data. Unlike Messrs. O’Hagan and Sorensen, Mr. DeSpautz is not a member of MDS Associates, Limited Partnership, and has no financial interest in the outcome of this case. Called as a witness for plaintiffs, Mr. DeSpautz gave lengthy and complex testimony concerning the ’902 specification’s use of ownship and target velocity. On cross-examination Mr. DeSpautz could not provide a clear answer to a question that goes to the heart of this litigation:

Q [By Defense Counsel] Is the target velocity signal required by Claim 1 ... an input to calculating the CPA vector signal that is defined later in Claim 1?

A [By Mr. DeSpautz] When we talk about the CPA and the parameters which we show in the 902 patent, one of them which is time to CPA, as well as the CPA, the course, the bearing at CPA, target velocity is definitely needed.

Q So then is it your testimony that a target velocity signal is required by Claim 1 for use in calculating the CPA vector signal of Claim 1?

A I go back to what I said before, and that is in defining the parameters of closest point of approach, you need a target position, and you need a target velocity to define some of the parameters that we put in there.

Q I am not sure if your answer to my question is yes or no or if it is not possible for you to answer yes or no.

A I don’t think it’s possible for me to answer the.question yes or no.

As the parties alleging patent infringement, plaintiffs bear the burden of proving their case by a preponderance of the evidence. While clearly a talented mathematician, Mr. DeSpautz simply could not overcome Mr. Pease’s clear and concise testimony concerning the inputs employed by the equations in the ’902 specification. Viewing Mr. DeS-pautz’s testimony in conjunction with Mr. Sorensen’s prior inconsistent testimony, the court does not find plaintiffs’ testimony concerning the inputs for Claim 1 in the ’902 specification credible.

The court therefore interprets claim 1 of the ’902 patent as determining CPA Range and CPA Bearing by using two target relative position fixes.

2. Comparison of claim to accused devices

The second step of the infringement analysis requires the court to compare the properly construed claim with the accused devices. Ekchian, 104 F.3d at 1302. Because Claim 1 of the ’902 patent is drafted in means-plus-function language, the court must do more than simply compare the functions of Claim 1 with the functions of the accused devices. In addition, if the court finds a commonality of functions, the court then must determine whether the accused devices employ structures identical or equivalent to the structures disclosed in the ’902 specification. Only if there is a commonality of function and of structure can the court find that the accused devices inflinge the ’902 patent: “In order to meet a means-plus-function limitation, an accused device must (1) perform the identical function recited in the means limitation and (2) perform that function using the structure disclosed in the specification or an equivalent structure.” Carroll Touch, Inc., 15 F.3d at 1578 (Fed.Cir.1993) (citing Valmont Indus., Inc. v. Reinke Mfg. Co., 983 F.2d 1039, 1042 (Fed.Cir.1993)); see King Instruments Corp v. Perego, 65 F.3d 941, 945-46 (Fed.Cir.1995), cert. denied, — U.S. -, 116 S.Ct. 1675, 134 L.Ed.2d 778 (1996); Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931, 933-34 (Fed.Cir.1987) (en banc).

The parties agree that the accused representative systems perform the same function as claimed in Claim 1 of the ’902 patent. Claim 1 of the ’902 patent describes a computerized collision avoidance system that calculates CPA Range and CPA Bearing by processing an ownship velocity signal, a target position signal, and a target velocity signal. Likewise, each of the accused representative systems uses an ownship velocity signal, a target position signal, and a target velocity signal to obtain a CPA vector.¹⁰

Having found an identity of functions, the court must now compare the structures used by the accused representative systems with the structures disclosed in the ’902 specification. Both Messrs. Pease and Sorensen testified that the accused representative systems process the ownship and target velocity signals in a manner similar to DX 8. By subtracting the ownship velocity vector from the target velocity vector, the computer in the accused representative systems can locate the relative motion line. Once the computer locates the relative motion line, it can predict CPA Range and CPA Bearing by dropping a perpendicular from ownship to the target.

The means used by the accused representative systems to calculate CPA data are not identical to the means disclosed in the ’902 specification. The ’902 specification does not describe a means for predicting CPA Range and CPA Bearing by processing ownship and target velocity input signals. Rather, the ’902 specification describes a means for predicting CPA Range and CPA Beating by processing two target position input signals. This difference in inputs results in a difference in structures. While the accused representative systems locate the relative motion line by subtracting the ownship velocity vector from the target velocity vector, the equations in the ’902 specification locate the relative motion line by connecting the two target position fixes. As Mr. Pease testified, the accused representative systems calculate CPA Range and CPA Bearing in the manner depicted in DX 8, while the ’902 specification discloses a means for calculating CPA data similar to DX 5. The court finds that the structures used by the accused representative systems are not identical to the structures disclosed in the ’902 specification.

In addition, the court finds that the means used by the accused representative systems are not equivalent to the means disclosed in the ’902 specification. The Federal Circuit has explained that the equivalency analysis required by 35 U.S.C. § 112, para. 6 is different than the equivalency analysis required by the doctrine of equivalents: which adds nothing of significance. In the context of section 112, however, an equivalent results from an insubstantial change which adds nothing of significance to the structure, material, or acts disclosed in the patent specification. A determination of section 112 equivalence does not involve the equitable tripartite test of the doctrine of equivalents. As this court has stated, “the sole question” under section 112 involves comparison of the structure in the accused device which performs the claimed function to the structure in the specification. D.M.I., 755 F.2d at 1575; Durango Assocs. Inc. v. Reflange, Inc., 843 F.2d 1349,1357, 6 USPQ2d 1290,1295 (Fed.Cir. 1988).

Section 112 and the doctrine of equivalents have something in common. The word “equivalent” in section 112 invokes the familiar concept of an insubstantial change

Valmont Indus., Inc., 983 F.2d at 1043.

The accused representative systems’ use of ownship and target velocity input signals, rather than two target position input signals, results in a significant difference in structure. Mr. Pease testified that the accused representative systems can process CPA Range and CPA Bearing almost instantaneously, assuming the target ship has already been entered into the system. The system disclosed in the ’902 specification cannot calculate CPA data as quickly. When the operator of the ’902 system instructs the computer to calculate CPA Range and CPA Bearing, the computer first must record the position of the target at two points in time. In order to ensure that the relative motion line is accurate, the computer must wait up to ten minutes between the first and second position fixes:

Q [By Defense Counsel] Can you provide some examples illustrating why the differences between the structures in the AEGIS system for performing its CPA function are substantially different from those in the 902 patent?

A [By Mr. Pease] Yes. In the AEGIS system, when the operator calls for a CPA solution, that is, designates a target and presses the necessary button to produce the CPA solution, the result appears substantially instantaneously. That is, the computational procedure takes a small fraction of a second to produce the result.

The computer in the 902 patent similarly has a computer which gives a very rapid result, so the computation time I will refer to in both cases as essentially instantaneous.

But, in the system described in the patent specification calls for a CPA result, the target is designated and then a position fix is taken. Then, some time must be allowed to elapse, so that the second fix is far enough away from the first fix, so that the difference is accurate enough to produce a relative motion line. This is typically in marine situations anywhere from two to ten minutes. It’s very common to use six minutes for slow moving vessels, because that’s a tenth of an hour.

Therefore, the result is that when you get a solution by the method indicated in the patent, it takes minutes to get an answer. In the case of the military system or the AEGIS system, that answer must be produced instantaneously and the inputs are chosen so that will happen.

Because the structures of the accused representative systems are neither identical nor equivalent to the structures disclosed in the ’902 patent, the court finds that the accused representative systems do not infringe the ’902 patent.

Messrs. Pease and Sorensen provided valuable contributions during the trial. Mr. Pease testified as defendant’s expert witness, while Mr. Sorensen testified as both a factual and expert witness, revisiting the stand as a rebuttal witness. Their combined testimony extended over a three-day period, thereby affording the court an excellent opportunity to assess the witnesses’ credibility; demean- or; expertise; and familiarity with the patent art, accused devises, and prior art; as well as the overall impact of their testimony. The court notes that Mr. Pease was previously involved in settlement negotiations which sought to resolve plaintiffs’ claim of infringement against Raytheon, Mr. Pease’s former employer. This involvement, however, did not detract from Mr. Pease’s ability to serve as an objective expert witness in the case at bar. Although both witnesses were knowledgeable and sincere, ultimately Mr. Pease proved to be the more convincing on the issues of infringement. Mr. Sorensen could not rebut either convincingly or effectively Mr. Pease’s analysis of Figure 2A of the ’902 patent and DX 5, DX 8 and DX 12. Essentially, defendant made out a case of non-infringement that plaintiffs could not rebut, although plaintiffs were charged with the burden of proving infringement in the first place.

II. Invalidity

Pursuant to 35 U.S.C. § 282 (1994), a patent is presumed valid. Given this presumption of validity, the party asserting invalidity bears the burdens of going forward and of persuasion and must prove invalidity by clear and convincing evidence. Checkpoint Sys., Inc. v. United States Int’l Trade Comm’n, 54 F.3d 756, 761 (Fed.Cir.1995). Clear and convincing evidence is a greater standard of proof than preponderance of the evidence but a lesser standard than beyond a reasonable doubt. Buildex Inc. v. Kason Indus., Inc., 849 F.2d 1461, 1463 (Fed.Cir.1988). “ ‘Clear and convincing’ evidence has been described as evidence which produces in the mind of the trier of fact an abiding conviction that the truth of a factual contention is ‘highly probable.’ ” Price v. Symsek, 988 F.2d 1187, 1191 (Fed.Cir.1993) (quoting Buildex, 849 F.2d at 1463).

1. Section 112 para. 2; invalidity for indefiniteness

Defendant argues that the ’902 patent is invalid for indefiniteness pursuant to the second paragraph of 35 U.S.C. § 112. 35 U.S.C. § 112 para. 2 states that “[t]he specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.” The test for definiteness is whether the claim in question “reasonably apprises those of skill in the art of its scope.” In re Warmerdam, 33 F.3d 1354, 1361 (Fed.Cir.1994). According to defendant, because the ’902 specification does not disclose a means for performing the claimed function, the patent is indefinite:

The Government does not assert that the specification is not enabling; one of ordinary skill in the art in 1969 could have written a variety of operable computer programs for performing the claimed function. The question is whether the ’902 specification identifies which of the many possible computer programs for performing the claimed function is foreclosed from future enterprise.

Def s Br. filed Oct. 4,1996, at 19-20.

The Federal Circuit recently addressed the applicability of the definiteness requirement to claims drafted in means-plus-function language:

Although paragraph six statutorily provides that one may use means-plus-funetion language in a claim, one is still subject to the requirement that a claim “particularly point out and distinctly claim” the invention. Therefore, if one employs means-plus-function language in a claim, one must set forth in the specification an adequate disclosure showing what is meant by that language. If an applicant fails to set forth an adequate disclosure, the applicant has in effect failed to particularly point out and distinctly claim the invention as required by the second paragraph of section 112.

In re Donaldson Co., 16 F.3d 1189, 1195 (Fed.Cir.1994) (en banc).¹¹

Cite as 37 Fed.

Cl. 611 (1997)

Applying the rule set forth in In re Donaldson Co., the courts finds Claim 1 of the ’902 patent invalid for indefiniteness. As discussed in section I of this opinion, Claim 1 of the ’902 patent claims an automated collision avoidance system that calculates CPA Range and CPA Bearing using an ownship velocity input signal, a target position input signal, and a target velocity input signal. The ’902 specification, however, does not disclose a means for calculating CPA Range and CPA Bearing using the above-mentioned input signals. Rather, the ’902 specification discloses a means for calculating CPA Range and CPA Bearing using two target position input signals.

The parties agree that one of ordinary skill in the art of automated collision avoidance radar systems, as of August 1969, would have been a marine collision avoidance radar design engineer “with the equivalent of at least a university level bachelor of science degree in electrical engineering and the equivalent of three years experience in the field.” Def s Br. filed Oct. 4,1996, at 35. The ’902 specification’s failure to disclose a means for the claimed function leaves one skilled in the art of automated collision avoidance radar systems unable to discern the scope of the patent. “[A] means clause does not cover every means for performing the specified function.” Laitram Corp. v. Rexnord, Inc., 939 F.2d 1533,1536 (Fed.Cir.1991) (emphasis omitted). Rather, a means clause is limited to the means disclosed in the specification and the equivalents thereof. Sofamor Danek Group, Inc. v. DePuy-Motech, Inc., 74 F.3d 1216, 1220 (Fed.Cir.1996). Without a disclosure of a means for processing ownship and target velocity input signals to calculate CPA data, one skilled in the art is left to guess which of the multiple means for performing the claimed function is covered by the patent. When one skilled in the art cannot discern

“what is meant by that [the means-plus-function] language,” the patent is indefinite. In re Donaldson Co., 16 F.3d at 1195.

2. Section 102(g); anticipation — prior invention

Defendant contends that the Navy invented the NTDS Model III system prior to plaintiffs’ invention of its collision avoidance system and that, therefore, the NTDS Model III system anticipated the ’902 patent. 35 U.S.C. § 102(g) (1994), provides in full:

A person shall be entitled to a patent unless—

(g) before the applicant’s invention thereof the invention was made in this country by another who had not abandoned, suppressed, or concealed it. In determining priority of invention there shall be considered not only the respective dates of conception and reduction to practice of the invention, but also the reasonable diligence of one who was first to conceive and last to reduce to practice, from a time prior to conception by the other.[¹²]

Section 102(g) recognizes the well-established principle that a patent right belongs to the first inventor, not the first person to file a patent application. In furtherance of this principle, section 102(g) awards priority to the “ ‘first party to reduce an invention to practice unless the other party can show that it was the first to conceive the invention and that it exercised reasonable diligence in later reducing that invention to practice.’” Mahurkar v. C.R. Bard Inc., 79 F.3d 1572, 1577 (Fed.Cir.1996) (quoting Price, 988 F.2d at 1190).

The NTDS Model III system is an earlier version of the NTDS Model 4.1 Baseline 7A for the FFG-7 Class Frigates (“NTDS FFG-7”), one of the accused systems in this case. By 1970 the NTDS Model III was operating on approximately 36 Navy ships. Like NTDS FFG-7, the NTDS Model III processes four input signals: ownship velocity, ownship position, target position, and target velocity. The means by which the NTDS Model III processes the input signals to calculate CPA Range and CPA Bearing are substantially the same as the means by which NTDS FFG-7 processes the same signals.

Plaintiffs advance two contentions against a finding that the NTDS Model III is a prior invention pursuant to 35 U.S.C. § 102(g). First, plaintiffs argue that defendant has failed to prove by clear and convincing evidence that the NTDS Model III contained each of the elements claimed by Claim 1 of the ’902 patent. Plaintiffs base this contention on the fact that the NTDS Model III technical documents produced by defendant during discovery were functional in nature and did not indicate how the system performed the CPA function. Second, plaintiffs argue that the NTDS Model III, by virtue of its classified status, was suppressed or concealed and therefore cannot invalidate the ’902 patent.

While both arguments raise interesting legal issues, they are not necessary for ruling on this defense. If the quality of defendant’s evidence is sufficient and if the Government did not suppress or conceal the NTDS Model III, defendant cannot succeed with its prior invention defense unless defendant identifies a single prior art reference that “discloses each and every limitation of the claim.” Glaxo Inc. v. Novopharm Ltd., 52 F.3d 1043, 1047 (Fed.Cir.), cert. denied, 116 S.Ct. 516 (1995). The Federal Circuit has phrased the test for anticipation as follows: “ ‘[T]hat which infringes if later anticipates if earlier.’ ” Polaroid Corp. v. Eastman Kodak Co., 789 F.2d 1556, 1573 (Fed. Cir.1986) (quoting Peters v. Active Mfg. Co., 129 U.S. 530, 537, 9 S.Ct. 389, 392, 32 L.Ed. 738 (1889)). Therefore, to determine whether the NTDS Model III anticipates the ’902 patent, the court must ask whether the NTDS Model III would have infringed the ’902 patent if developed after the ’902 invention. Donald S. Chisum, Patents § 3.02[1] (1996).

The NTDS Model III would not have infringed the ’902 patent if developed after the latter’s invention. As defendant itself proved on the issue of infringement, the NTDS FFG-7 does not employ the same or equivalent means for calculating CPA Range and CPA Bearing as the equations disclosed in the ’902 specification. Similarly, defendant proved that the NTDS Model III employs substantially similar means as the NTDS FFG-7. If the NTDS FFG-7 does not infringe the ’902 patent, then, given the similarity between the NTDS FFG-7 and the NTDS Model III, the NTDS Model III does not employ the same elements as Claim 1 of the ’902 patent and therefore does not anticipate the ’902 patent. Having failed to prove an identity of elements, defendant cannot succeed on its section 102(g) prior invention defense.¹³

Although the lack of identity between the NTDS Model III and Claim 1 of the ’902 patent disposes of defendant’s section 102(g) defense, in order to provide a complete record on the issue, the court considers a related issue concerning section 102(g) strenuously litigated by the parties at trial — whether the NTDS Model III was suppressed or concealed. Plaintiffs argue that the Navy’s classification of the NTDS Model III rendered the prior art suppressed or concealed because the public did not have access to the invention. Defendant counters that the act of classification does not remove the NTDS Model III from the public.

A defense of prior invention will fail if the prior art was “abandoned, suppressed, or concealed.” 35 U.S.C. § 102(g). A precise definition for this phrase does not exist. As Judge Rich explained in Paulik v. Rizkalla, 760 F.2d 1270, 1279 (Fed.Cir.1985) (en banc) (Rich, J. concurring), the words “suppressed or concealed” were included in section 102(g) to codify preexisting case law relating to priority issues in interference and infringement suits. Therefore, to understand the meaning of suppression and concealment, one must consult the cases upon which section 102(g) is based. Foremost among these cases is Mason v. Hepburn, 13 App.D.C. 86 (1898):

The true ground of the doctrine, we apprehend, lies in the policy and spirit of the patent laws and in the nature of the equity that arises in favor of him who gives the public the benefit of the knowledge of his invention, who expends his time, labor, and money in discovering, perfecting, and patenting, in perfect good faith, that which he and all others have been led to believe has never been discovered, by reason of the indifference, supineness, or willful act of one who may, in fact, have discovered it long before.

Mason, 13 App.D.C. at 96.

More recently, the Federal Circuit reiterated the policy underlying the suppressed or concealed proviso of section 102(g): “[I]t ... encourages prompt public disclosure of an invention by penalizing the unexcused delay or failure of a first inventor to share the ‘benefit of the knowledge of [the] invention’ with the public after the invention has been completed.” Checkpoint Sys., 54 F.3d at 761 (quoting Paulik, 760 F.2d at 1280). In Lutzker v. Plet, 843 F.2d 1364, 1366 (Fed.Cir. 1988), the court described the different ways by which an inventor can disclose an invention to the public: 1) file a patent application; 2) describe the invention in a publicly disseminated document; and 3) use the invention publicly.

The Federal Circuit or its predecessor (the “C.C.P.A.”) has concluded that an invention was suppressed or concealed where the finding was consistent with the policy considerations underlying section 102(g). For example, the C.C.P.A. held an invention suppressed and concealed where the inventor sat idly on the completed invention, only becoming motivated when he sensed another inventor creating a similar invention. See Brokaw v. Vogel, 429 F.2d 476, 480 (C.C.P.A. 1970). Like-wise, the C.C.P.A. held an invention suppressed and concealed where the inventor kept his invention secret while at the same time using the invention to market a commercially valuable product. See Palmer v. Dudzik, 481 F.2d 1377, 1387 (C.C.P.A. 1973).

Unfortunately, applying the 102(g) policy considerations to the facts of the case at bar is not a simple task. The court must balance two competing public policies, both of which operate in the name of the public good. On the one hand, the patent system encourages and demands public disclosure of inventions. Plaintiffs are inventors who went about their creative process with no knowledge of the Navy’s NTDS Model III system. Plaintiffs invested large amounts of time, energy, and money inventing a product which they had no reason to believe had been invented before, ie., there was no patent application, no dissemination of a public document, and no public use of the Navy’s NTDS Model III system. From this perspective, section 102(g) appears to warrant a finding of suppression and concealment.

The other perspective reveals that the Navy developed a command data system that was deployed in at least 36 navy ships during a time of increased global unrest. Although the Navy classified its invention, the NTDS Model III system undoubtedly contributed to the defense of the nation. The question thus arises whether the Navy’s classification of the NTDS Model III system violates the public disclosure requirement of Mason and its progeny, or whether defendant can satisfy the public disclosure requirement even though the public’s access to the Navy’s invention was limited by law.

In Del Mar Eng’g Lab. v. United States, 207 Ct.Cl. 815, 524 F.2d 1178 (1975), the Court of Claims addressed a similar issue. Plaintiff invented and patented a target structure which simulated a self-propelled aircraft for purposes of testing missile control systems. It sued claiming that the Navy was infringing its patent. Defendant rejoined that the Navy was the prior inventor of the same concept and that plaintiffs patent therefore was invalid under section 102(g). Plaintiff argued that, given the classified status of defendant’s prior art, the prior art was suppressed or concealed. In weighing the competing policy choices, the Court of Claims stated:

It is difficult to view the secrecy imposed on work by a security classification as being hostile to the public good. To the contrary, presumably it is for the public good, as perceived by those in responsible positions of government, that certain types and kinds of work are at one time or another placed in a classified status in which a security clearance is required before anyone is permitted to have access to it. But even then, it frequently is the case that the work itself is intended to be put into use for the public benefit. In other words, the classification restriction may not suppress the work itself, only access to that work. For example, a classified weapon may be manufactured in quantity and placed in the nation’s arsenal of weapons, all for the benefit of the public, but yet be kept secret both from the public at large and those working in that same art. In such circumstances, is the Government, which imposed the secrecy order for the public good, obliged to pay compensation to a second inventor who, working independently and without knowledge of the classified article, makes the same invention, but at a later time? Because of considerations such as these, it is believed that the fact of security classification should not be regarded per se as a suppression and concealment; rather, it should be viewed as but one fact in the totality of particular facts applicable to the specific situation under consideration---In the absence of any showing that defendant, by its security classification system, attempted to exclude the public from the benefit of this work, it is considered not to have been suppressed or concealed for purposes of 35 U.S.C. § 102(g).

207 Ct.Cl. at 826-27, 524 F.2d at 1185 (emphasis added).

Since Del Mar only one other case appears to have involved section 102(g) and a classified military prior art system. In Siemens Aktiengesellschaft v. United States, 26 Cl.Ct. 980 (1992), plaintiff was the owner by assignment of a patent for traveling wave tubes. In response to plaintiffs claim of infringement, defendant argued that plaintiffs patent was invalidated by a prior invention. The prior invention was conceived by a third party, who applied for a patent approximately 45 days after plaintiffs patent application. As in Del Mar, the prior invention was classified, and plaintiff sought a finding of suppression and concealment.

The court restated the basic holding of Del Mar:

As an initial matter, defendant accurately asserts that the “SECRET” classification on the Navy proposals does not establish per se suppression or concealment.... [A] security classification is merely one fact to consider in determining whether an invention was suppressed or concealed. Unless plaintiff can show that defendant attempted to exclude the invention from the public through its security system, the “Secret” classification on the Navy proposal is insufficient evidence in itself of suppression or concealment.

26 Cl.Ct. at 987. Then, reviewing the facts of the case, the court concluded that sufficient steps had been taken to exclude the prior art from the public:

Plaintiff, however, contends that through a deliberate and complicated classification system, the Noland gun was excluded from the public. The cover page of the proposal Hughes made to the Navy which ultimately resulted in the ’0205 contract contains two “SECRET” stamps, two “CONFIDENTIAL” stamps, a “SECRET” classification, a “NATIONAL SECURITY INFORMATION” stamp, and the following warnings:

This document contains information affecting the national defense of the United States, within the meaning of the Espionage Laws, Title 18, U.S.C., Sections 793 and 794, the transmission or revelation of which in any manner to an unauthorized person is prohibited by law.

The court finds it difficult to argue with plaintiffs assertion that, taken as a whole, these warnings were sufficient to exclude the Noland Gun from the public.

26 Cl.Ct. at 987.

The court in Siemens conflated the concepts of public benefit and public access. Del Mar, the binding precedent, separates them and teaches that the mere fact of classification does not lead to a per se finding of suppression and concealment because the classified prior art may nonetheless yield a public benefit. However, Del Mar also instructs that the act of classification will result in a finding of suppression and concealment where the Government uses the classification system to exclude the public from the benefit of its invention.

The NTDS Model III system provided numerous benefits to the public. First, the system equipped the military with computerized anti-air warfare capabilities far superior to those formerly provided by manual radar tracking technology. Second, the system allowed multiple ships to share digital data, thus increasing the military’s ability to perform coordinated combat direction and decision making. Third, the development and continued maintenance of the NTDS Model III system spawned hundreds of civilian jobs in the United States. Fourth, the sale of the NTDS Model III system to West Germany and France permitted NATO allies to adopt the state-of-the-art defense systems used by the United States military.

Although the NTDS Model III provided a demonstrable public benefit, the classification system put in place for the NTDS Model III was a successful attempt to exclude the public from access to the prior art system. The parties stipulated that all of the NTDS Model III technical documentation were classified at at least the “confidential” level. Further, August Henry Giobbi, defendant’s fact witness for the NTDS Model III, testified that access to technical documentation was limited to those with a proper security clearance and a need to know:

Q (By Plaintiffs’ Counsel) Mr. Giobbi, while you were in the Navy working on the NTDS Model III system, who would have had access to the classified portions of the documents that we reviewed today?

A It would be properly cleared contractor personnel from Sperry Univae, Computer Sciences Corporation and the Navy personnel.

Q And so the Navy personnel would have had to have proper clearance as well?

A As you know, it’s on a need to know basis, but yes.

Q And what da you mean by need to know basis?

A The military doesn’t just give access to documents unless you have a need to know, as they say, which means you are involved with the process.

Q Mr. Giobbi, were these documents accessible to the public?

A No.

In addition to being classified, all of the NTDS Model III technical documents introduced by defendant contained the same warning as the technical documents in Siemens. One of these documents, DX 93, list- • ed an additional warning: “NATIONAL SECURITY INFORMATION Unauthorized Disclosure Subject to Criminal Sanctions.”

If the analysis of this defense had proceeded to this point, the court would balance the fact of classification against the benefits to the public. On balance the court would find that defendant had failed to carry its burden of clear and convincing evidence because the considerations are in equipoise.

3. Section 102(b) — on-sale bar

Defendant also argues that the ’902 patent is invalid under the on-sale bar of 35 U.S.C. § 102(b). In support of this invalidity claim, defendant points out that the Navy sold the SATIR system to West Germany more than one year before plaintiffs filed their patent application.

In 1966 the Navy offered to sell to the Federal Republic of Germany three command data systems known as SATIR. Due to the inability of the German government to work directly with a contractor, the Navy established a two-step procedure for the sale. Initially, the Navy purchased the SATIR systems from the Sperry Rand Corporation (Univac division) (“Sperry”). Later the Navy resold the SATIR systems to Germany. The final step of the transaction took place on January 8-9, 1969, when Germany accepted the systems aboard ship in Bath, Maine.

The SATIR systems sold to Germany were modified NTDS Model III systems. Sperry customized the systems to meet the space limitations and weapons systems of the German ships on which the systems were to be installed. These systems performed the same CPA functions in essentially the same manner as the non-customized NTDS Model III system. Given the national security implications of this sale, the SATIR program design specifications and operators’ manuals were classified “confidential.”¹⁴ However, the sales documentation was not classified.

Section 102(b) provides:

A person shall be entitled to a patent unless—

(b) the invention was patented or described in a printed publication in this or a foreign country or in public use or on sale in this country, more than one year prior to the date of the application for patent in the United States.

A party asserting the on-sale bar must prove that 1) a sale or offer to sell was made; 2) in the United States; 3) more than one year before the application for the patent the movant seeks to invalidate was filed; and 4) “the subject matter of the sale or offer to sell fully anticipated the claimed invention or would have rendered the claimed invention obvious by its addition to the prior art” UMC Elecs. Co. v. United States, 816 F.2d 647, 656 (Fed. Cir.1987). Plaintiffs present three reasons why the on-sale bar should not invalidate the ’902 patent. First, as with the NTDS Model III system, plaintiffs argue that the SATIR technical documents produced by defendant during discovery were functional in nature, so that defendant cannot prove that the SATIR system contained each of the elements claimed by Claim 1 of the ’902 patent. Second, plaintiffs suggest that the SATIR system was not “on sale in this country,” as required by section 102(b). Third, plaintiffs contend that Congress did not intend for the on-sale bar to apply where the prior art is classified and the sale does not make the invention freely available to the general public. Ruling on the applicability of the on-sale bar does not require a detailed discussion of all of plaintiffs’ arguments.

The on-sale bar does not require a strict identity of elements, as do the novelty provisions of section 102. See UMC Elecs. Co., 816 F.2d at 656. The sale of a device meeting all of the other requirements of the on-sale bar will invalidate a later patent if it “would have rendered the claimed invention obvious by its addition to the prior art.” Id. In the case at bar, however, defendant does not contend that the SATIR system renders Claim 1 of the ’902 patent obvious, although defendant argues that the SATIR system anticipated Claim 1 of the ’902 patent.

At trial defendant proved that the NTDS FFG-7 does not infringe Claim 1 of the ’902 patent because the NTDS FFG-7 employs different means than the equations in the ’902 specification. While the NTDS FFG-7 uses ownship and target velocity input signals to calculate CPA Range and CPA Bearing, the equations in the ’902 specification use two target position input signals. Defendant also proved that the SATIR system employs the same means as NTDS FFG-7 for calculating CPA Range and Bearing. As discussed previously, the Federal Circuit’s test for anticipation states: “ ‘[T]hat which infringes if later anticipates if earlier.’ ” Polaroid Corp., 789 F.2d at 1573 (quoting Peters, 129 U.S. at 537, 9 S.Ct. at 392). Given the identity between NTDS FFG-7 and the SATIR system, the SATIR system would not have infringed Claim 1 of the ’902 patent if it had been developed after, rather than before, the invention of the ’902 system. Accordingly, the SATIR system does not anticipate Claim 1 of the ’902 patent, and the on-sale bar does not apply to the SATIR transaction.¹⁵

Because defendant failed to prove an exact identity between the SATIR system and Claim 1 of the ’902 patent, defendant’s on-sale bar defense must fail. However, for completeness the court will discuss the two other on-sale bar issues raised by the parties.

Plaintiffs argue that the SATIR system was not “on sale in this country” because the system allegedly was not accepted in this country. In making this argument, plaintiffs do not indicate whether they are referring to the Sperry-Navy sale or Navy-Germany sale. The court discusses each sale in turn.

Defendant called Captain Alfred McLean Bettis, who was in charge of the Navy office that wrote the technical specifications for the SATIR system. He testified that all of the steps in the Sperry-Navy sale took place in the United States. Plaintiffs did not elicit any testimony that would diminish that of Captain Bettis or come forward with any evidence to the contrary. Given Captain Bettis’ persuasive testimony, the court finds that the Sperry-Navy sale took place in the United States.

Concerning the Navy-Germany sale, a “product is ‘on sale’ in the United States, within the proscription of the statute, if substantial activity prefatory to a sale occurs in the United States.” Robbins Co. v. Lawrence Mfg. Co., 482 F.2d 426, 434 (9th Cir. 1973). Captain Bettis testified that pre-con-tract negotiations, contract amendment negotiations, construction of components, integration of the systems, functional testing, installation aboard ship, shipboard testing, and acceptance all took place in the United States. Again, plaintiffs offered no evidence to the contrary. As a result, based on Captain Bettis’ testimony, the court concludes that the Navy-Germany sale took place “in this country” for purposes of the section 102(b) on-sale bar.

Plaintiffs also contend that the on-sale bar is inapplicable to this case because the SATIR system and its technical documentation were classified and thus were not freely available to the public. According to plaintiffs, the policy considerations underlying the on-sale bar provision of section 102(b) are not implicated by a third-party secret sale. Defendant counters that the secret nature of the sales has no bearing on the application of the on-sale bar.

“[W]hether a device has been placed on sale is not subject to a mechanical rule. On the contrary, the on-sale determination depends on the totality of the circumstances, considered in view of the policies underlying section 102(b).” In re Mahurkar Double Lumen Hemodialysis Catheter Patent Litig., 71 F.3d 1573, 1577 (Fed.Cir.1995) (citations omitted). There are four policies underlying the section 102(b) on-sale bar:

(1) A policy against removing inventions from the public which the public has justifiably come to believe are freely available to all as a consequence of prolonged sales activity.

(2) A policy favoring prompt and widespread disclosure of new inventions to the public. The inventor is forced to file promptly or risk possible forfeiture of his patent rights due to prior sales.

(3) A policy to prevent the inventor from commercially exploiting the exclusivity of Ms Mvention beyond the statutorily authorized 17 year period.

(4) A policy to give the mventor a reasonable amount of time-following sales activity (set by statute as one year) to determine whether a patent is a worthwhile rnvestment. TMs benefits the public because it tends to minimize the fifing of inventions of only margmal public mterest.

Siemens, 26 Cl.Ct. at 989 (citing UMC Elecs Co., 816 F.2d at 652); see In re Mahurkar, 71 F.3d at 1577.

To date, the Federal Circuit has not addressed the precise issue raised by plaintiffs — whether the third-party sale of a device subject to military classification triggers the on-sale bar. The Federal Circuit, however, has offered guidance in dicta, wMch the court considers in conjunction with the policies underlying the on-sale bar.

As a general rule, the on-sale bar applies to sales made by third parties, as well as to sales made by the mventor. In re Epstein, 32 F.3d 1559, 1564 (Fed.Cir.1994). “An exception to this general rule exists where a patented method is kept secret and remains secret after a sale of the unpatented product of the method. Such a sale prior to the critical date is a bar if engaged in by the patentee or patent applicant, but not if engaged m by another.” In re Caveney, 761 F.2d 671, 675 (Fed.Cir.1985) (citing W.L. Gore & Assocs., Inc. v. Garlock, Inc., 721 F.2d 1540, 1550 (Fed.CM.1983)). The Federal CMemt created the tMrd-party-secretmethod exception because the sale of a product, where the method is kept secret, prevents the public from learning of the invention. Specifically, the Federal CMcuit stated:

Early public disclosure is a linehpm of the patent system. As between a prior mventor who benefits from a process by selling its product but suppresses, conceals, or otherwise keeps the process from the public, and a later inventor who promptly files a patent application from which the public will gain a disclosure of the process, the law favors the latter.

W.L. Gore & Assocs., Inc., 721 F.2d at 1550.

The SATIR sales disclosed the invention to the purchaser, so the tMrd-party-secret-method exception offers no solace to plaintiffs. However, in In re Caveney, the Federal CMcuit indicated, in dicta, that the on-sale bar does not apply to thMd-party sales that, while disclosing the invention to the purchaser, do not make the invention accessible to the public: “The ‘on sale’ provision of 35 U.S.C. § 102(b) is directed at precluding an mventor from commercializing his invention for over a year before he files his application. Sales or offers made by others and disclosing the claimed invention implicate the ‘public use’ provision of 35 U.S.C. § 102(b).” 761 F.2d at 675 n. 5.

In re Caveney does not provide guidance as to how a trial court should incorporate the section 102(b) public use provision into its analysis of a thMd-party sale. The trial court, after finding that a sale has occurred, could apply, as an additional element, precedent governing the public use provision. Under this approach Germany’s use of the SA-TIR systems does not constitute a section 102(b) public use. See Del Mar Eng’g, 207 Ct.Cl. at 822, 524 F.2d at 1182 (holding that thMd-party use of classified invention does not constitute public use); W.L. Gore, 721 F.2d at 1549-50 (holding that use of invention, where use was restricted by secrecy agreement, does not constitute public use).

Alternatively, the trial court could inquMe whether the sale in question made the prior art accessible to the public. Requiring a court to consider the question of public accessibility when faced with a third-party disclosing sale comports with the policies underlying the statute. As the Federal CMcuit has explained, not all of the policies underlying section 102(b) are implicated by a thMd-party sale. For example, the policy proMbiting prolonged exclusive exploitation of an invention and the policy allowing an inventor a reasonable time to judge the utility of an mvention do not come into play when the sale is made by a thMd party. General Elec. Co., 228 Ct.Cl. at 202-03, 654 F.2d at 61-62. Accordingly, to apply the on-sale bar to a thMd-party sale, the court must find that either the policy against removing inventions from the public or the policy promoting early disclosure of new inventions is furthered.¹⁶ Both of these policies focus on public accessibility, not simply the fact that a sale was made.¹⁷

Holding that the on-sale bar invalidates the ’902 patent serves neither of the two above policies. First, given that the SATIR technical documentation was classified “confidential”, the public would not have “justifiably come to believe” that the SATIR system was “freely available to all as a consequence of prolonged sales activities.” Although the fact of the sale itself was not classified, the public could not know the contents of the sale. In addition to the classification system imposed by the Navy, the Navy-Germany contract prohibited Germany from reselling the SATIR systems and required Germany to employ a classification system equivalent to that employed by the United States. Second, the policy encouraging prompt public disclosure of new inventions is not furthered by invalidating plaintiffs’ invention. Because the SATIR technical information was classified, the inventors of the system covered by the ’902 patent could not have known, either actually or constructively, the contents of the SATIR sale. Furthermore, plaintiffs filed a disclosure document with the PTO on August 26, 1969, at least one year before they had built the first working prototype. Based on the fact that the SATIR sales do not implicate any of the policies underlying section 102(b), the court finds that the on-sale bar does not apply to the facts of this case.¹⁸

CONCLUSION

Accordingly, based on the foregoing, the Clerk of the Court shall enter judgment for defendant.

No costs.

1 . The court acknowledges the parties' substantial efforts to develop this protocol, which the court suggested and ultimately required. Trial confirmed that the protocol streamlined the proofs and did not disserve the full presentation of the parties’ respective cases. See RCFCl(a)(2).

2 . Messrs. O’Hagan, Sorensen, and DeSpautz testified at trial. Excerpts of Mr. Carroll’s deposition were admitted.

3 . Defendant asserts that plaintiffs’ infringement claims accruing before October 29, 1987, the date on which plaintiffs filed their administrative claim against the Navy, are barred by the doctrine of laches. Defendant bases this argument on an undated letter sent by Mr. O'Hagan to Mr. Bemis, warning the latter that the ’902 patent was being infringed by “the U.S. Gov’t in their radar (military) systems." Defendant surmises that the letter was written and sent “by later 1979 or early 1980,” and the evidence supports a finding to that effect. Plaintiffs respond that Mr. O’Hagan, when writing the letter, really did not actually know that defendant was infringing the patent.

To succeed on its laches defense, defendant has the burden of proving two elements: "1. [T]he plaintiff delayed filing suit for an unreasonable and inexcusable length of time from the time the plaintiff knew or reasonably should have known of its claim against the defendant, and 2. the delay operated to the prejudice or injury of the defendant.” A.C. Aukerman Co. v. R.L. Chaides Const. Co., 960 F.2d 1020, 1032 (Fed. Cir.1992) (quoting Costello v. United States, 365 U.S. 265, 282, 81 S.Ct. 534, 542, 5 L.Ed.2d 551 (1961)). If the court finds the period of delay to be more than six years, prejudice is presumed, and the burden of proving excuse for the delay and lack of prejudice shifts to the patentee. Hall v. Aqua Queen Mfg., Inc., 93 F.3d 1548, 1553 (Fed.Cir.1996). Because the accused systems at least were partially classified, it cannot be said that plaintiffs knew or should have known that they had viable infringement claims against the Navy and Coast Guard in 1980.

4 . The parties have labeled each of the four limitations of Claim 1 with an alphabetical label. When referring to the individual limitations, the court uses the letters (a)-(d) as shown in the quoted provisions.

5 . Defendant offers two different, but related, arguments concerning the construction of the '902 patent. First, defendant argues that the equations in the '902 specification do not identify a means for predicting CPA Bearing. Second, defendant argues that the equations for predicting CPA Range in the '902 specification employ different inputs than those claimed by Claim 1. In response to defendant's first argument, plaintiffs respond that the equations for predicting CPA Range also yield CPA Bearing data. Because the court agrees with defendant's second argument, it is not necessary to address the first argument independently. If plaintiffs are correct that the equations for predicting CPA Range also predict CPA Bearing, those equations employ two target positions as inputs, not ownship and target velocity.

6 . Defendant also argues that plaintiffs’ failure to disclose a means for performing the claimed function renders the '902 patent invalid for indefiniteness. The court addresses this argument in section II infra.

7 . Mr. Pease holds patents for each of above-mentioned Raytheon collision avoidance systems.

8 . Because he is a partner in plaintiff MDS Associates, Limited Partnership, Mr. Sorensen has a large financial interest in this case, a fact which, in combination with his inconsistent testimony, must be noted.

9 . Mr. Sorensen took issue with DX 12 and, during his direct examination, drew a diagram depicting his understanding of the inputs employed by the '902 specification to predict CPA. See PX 614B. Mr. Sorensen’s diagram indicates that the equations in the ’902 specification employ ownship and target velocity input signals. Once again, Mr. Sorensen's testimony and diagram were undermined by prior inconsistent testimony given at an earlier deposition. At Mr. Sorensen's October 20, 1995 deposition, defense counsel presented him with a diagram showing ownship velocity, target position, and target velocity as input signals. See DX 38. Mr. Sorensen testified that the diagram was not representative of the inputs in the '902 specification because the '902 specification did not employ target velocity to predict CPA.

10 . Two of the accused representative systems, NTDS and AEGIS, sometimes employ ownship position as a fourth input signal. The use of ownship position as an input occurs when NTDS and AEGIS compute CPA data between two ships other than ownship. When NTDS and AEGIS compute CPA between ownship and a target, these systems do not employ an ownship position signal because ownship’s position is by definition zero. HYCATS, COMDAC, and AN/ SPA-25G are not capable of computing CPA between two ship's other than ownship, and thus they always employ three input signals.

11 . The holding in In re Donaldson Co. follows logically from the policy considerations addressed by the enactment of 35 U.S.C. § 112 para. 6. Prior to the enactment of this provision, the Supreme Court prohibited the use of means-plus-function language, fearing that such language was overbroad and ambiguous. See Halliburton Oil Well Cementing Co. v. Walker, 329 U.S. 1, 9, 67 S.Ct. 6, 10, 91 L.Ed. 3 (1946); see also Valmont Indus., Inc., 983 F.2d at 1042 (discussing Supreme Court’s concern with means-plus-function language). In 1952

Congress decided to permit broad means-plus-function language, but provided a standard to make the broad claim language more definite.... Section 112 thus permits means-plus-function language in a combination claim, but with a "string attached.” The "attached string" limits the applicant to the structure, material, or acts in the specification and their equivalents.

Valmont Indus., Inc., 983 F.2d at 1042. By limiting a claim drafted in means-plus-function language to the means and equivalents thereof disclosed in the specification, 35 U.S.C. § 112 para. 6 responds to and addresses the fear of indefiniteness expressed by the Supreme Court in Halliburton Oil Well. Valmont Indus., Inc., 983 F.2d at 1042 ("A claim limitation described as a means for performing a function, if read literally, could encompass any conceivable means for performing the function. This second clause [of 35 U.S.C. § 112 para. 6] confines the breadth of protection otherwise permitted by the first clause.”). A patentee who fails to disclose a means for the claimed function has in effect ignored the “string attached" by 35 U.S.C. § 112 para. 6 and has made it impossible for one skilled in the art to understand the breadth of the claim, thus creating the very problem that prompted the Supreme Court to prohibit the use of functional language in Halliburton Oil Well.

12 . Section 102(g) is usually invoked in the patent interference context. However, a party may also raise section 102(g) as a defense in a patent infringement suit. Checkpoint Sys., 54 F.3d at 761.

13 . Because the court finds that the NTDS Model III does not anticipate Claim 1 of the '902 patent, it is unnecessary to address plaintiffs’ argument concerning the quality of the NTDS Model III technical documentation produced during discovery.

14 . DX 140, the Department of Defense form memorializing the offer and acceptance of the SA-TIR sale between the Navy and Germany, contained the following clause:

[The purchaser] [s]hall not transfer title to, or possession of, the items furnished under this sales agreement to any person, or organization, or other government, unless the consent of the Government of the United States has first been obtained. It shall not disclose, dispose of, or permit use of any plans, specifications or information pertaining to any items furnished in connection with this transaction, except to the extent authorized by the Government of the United States. To the extent that any items, plans, specifications, or information furnished in connection with this transaction may be classified by the Government of the United States for security purposes, the Purchaser shall maintain a similar classification and employ all measures necessary to preserve such security, equivalent to those employed by the Government of the United States, throughout the period during which the Government of the United States may maintain such classification. The Government of the United States will notify the purchaser if the classification is changed. The Purchaser will ensure, by all means available to it, respect for the proprietary rights in any defense article and any plans, specifications, or information furnished, whether patented or not.

15 . Because the court finds that the SATIR system does not anticipate Claim 1 of the '902 patent, it is unnecessary to address plaintiffs’ argument concerning the quality of the SATIR technical documentation produced during discovery.

16 . In General Elec. Co., the Court of Claims stated:

Where the sale is by one other than the inventor ... it would seem that the policy against extended commercial exploitation and the policy favoring the filing of only worthwhile inventions could be said not to apply. Nevertheless, it is well established that a placing of the invention "on sale” by an unrelated third party more than 1 year prior to the filing of an application for patent by another has the effect under § 102(b) of invalidating a patent directed to that invention. Accordingly, Congress should be held to have concluded, at the least, that the policy against removing inventions from the public domain and the policy favoring early patent filing are of sufficient importance in and of themselves to invalidate a patent where the invention is sold by one other than the inventor or one under his control. 228 Ct.Cl. at 202-03, 654 F.2d at 61-62 (citation omitted).

17 . As Professor Chisum has explained:

The general purpose behind all the bars is to require inventors to assert with due diligence their right to a patent through the filing and prosecution of a patent application. Lack of diligence by the inventor delays the time when the valuable invention is disclosed to the public through a published patent and delays the time when the invention will be free of the statutory period of monopoly.

Chisum, supra, § 6.01.

18 . Defendant also contends that the invention claimed by the '902 patent is obvious pursuant to 35 U.S.C. § 103. Because the court has considered four of defendant’s defenses and has concluded that the '902 patent is invalid for indefiniteness, it is not necessary to rule on this fifth defense.